Catálogo de publicaciones - libros

In this chapter, an introduction on the use of evolutionary computing techniques, which are considered as global optimization and search techniques inspired from biological evolutions, in the domain of system design is presented. A variety of evolutionary computing techniques are first explained, and the motivations of using evolutionary computing techniques in tackling system design tasks are then discussed. In addition, a number of successful applications of evolutionary computing to system design tasks are described.

In recent years, the use of hybrid Soft Computing methods has shown that in various applications the synergism of several techniques is superior to a single technique. For example, the use of a neural fuzzy system and an evolutionary fuzzy system hybridises the approximate reasoning mechanism of fuzzy systems with the learning capabilities of neural networks and evolutionary algorithms. Evolutionary neural systems hybridise the neurocomputing approach with the solution-searching ability of evolutionary computing. Such hybrid methodologies retain limitations that can be overcome with full integration of the three basic Soft Computing paradigms, and this leads to evolutionary neural fuzzy systems. The objective of this chapter is to provide an account of hybrid Soft Computing systems, with special attention to the combined use of evolutionary algorithms and neural networks in order to endow fuzzy systems with learning and adaptive capabilities. After an introduction to basic Soft Computing paradigms, the various forms of hybridisation are considered, which results in evolutionary neural fuzzy systems. The chapter also describes a particular approach that jointly uses neural learning and genetic optimisation to learn a fuzzy model from the given data and to optimise it for accuracy and interpretability.

The present chapter deals with the issues related to the evolution of optimal fuzzy logic controllers (FLC) by proper tuning of its knowledge base (KB), using different tools, such as least-square techniques, genetic algorithms, backpropagation (steepest descent) algorithm, ant-colony optimization, reinforcement learning, Tabu search, Taguchi method and simulated annealing. The selection of a particular tool for the evolution of the FLC, generally depends on the application. Some of the applications have also been included in this chapter.

This chapter presents a novel methodology for the customization of neural network based multi-classifiers used for the recognition of promoter regions in genomic DNA. We present a framework that utilizes genetic algorithms (GA’s) for the determination of optimal neural network parameters for better promoter recognition. The framework also presents a GA based method for the combination of the designed neural networks into the multi-classifier system.

The widespread deployment of WDM optical networks posts lots of new challenges to network designers. Traffic grooming is one of the most common and interesting problems. Efficient grooming of traffic can effectively reduce the overall cost of the network. But unfortunately, it has been shown to be NP-hard. Therefore, new heuristics must be devised to tackle them. Among those approaches, metaheuristics are probably the most promising ones. In this chapter, we present a thorough and comprehensive discussion on various metaheuristic approaches to the grooming of traffic in both static and dynamic patterns in WDM optical networks. Some future challenges and research directions are also discussed in this chapter.

This chapter presents EPSO (Evolutionary Particle Swarm Optimization), as an evolutionary meta-heuristic that implements a scheme of self-adaptive recombination, borrowing the movement rule from PSO (Particle Swarm Optimization). Besides the basic model, it discusses a Stochastic Star topology for the communication among particles and presents a variant called differential EPSO or dEPSO. The chapter presents results in a didactic Unit Commitment/Generator Scheduling Power System problem and results of a competition among algorithms in an intelligent agent platform for Energy Retail Market simulation where EPSO comes out as the winner algorithm.

Increasingly, research in the field of fuzzy theory is focusing on fuzzy sets (FSs) whose membership functions are themselves fuzzy. The key concept of such type-2 FSs is the footprint of uncertainty. It provides an extra mathematical dimension that equips type-2 fuzzy logic systems (FLSs) with the potential to outperform conventional (type-1) FLSs. While a type-2 FLS has the capability to model more complex relationships, the output of a type-2 fuzzy inference engine is a type-2 FS that needs to be type-reduced before defuzzification can be performed. Unfortunately, type-reduction is usually achieved using the computationally intensive Karnik-Mendel iterative algorithm. In order for type-2 FLSs to be useful for real-time applications, the computational burden of type-reduction needs to be relieved. This work aims at designing computationally efficient type-reducers using a genetic algorithm (GA). The proposed type-reducer is based on the concept known as (ET1FSs), a collection of type-1 FSs that replicates the input-output relationship of a type-2 FLS. By replacing a type-2 FS with a collection of ET1FSs, the type-reduction process then simplifies to deciding which ET1FS to employ in a particular situation. The strategy for selecting the ET1FS is evolved by a GA. Results are presented to demonstrate that the proposed type-reducing algorithm has lower computational cost and may provide better performance than FLSs that employ existing type-reducers.

The gyro-mirror line-of-sight stabilization platform used to maintain the line-of-sight of electro-optical sensors mounted on moving vehicles is a multivariate and highly nonlinear system. The system is also characterized by a peculiar phenomenon in which a movement about one axis will trigger off a coupled movement in the other axis. Furthermore, uncertainties such as noise, practical imperfections and additional dynamics are often omitted from the mathematical model of the system thus resulting in a non-trivial control problem. In order to handle the complex dynamics of the gyro-mirror as well as to optimize the various conflicting control objectives, a multi-objective artificial immune system framework which combines the global search ability of evolutionary algorithms and immune learning of artificial immune systems is proposed in this chapter for the design of the gyroscope recurrent neural network controller. In addition, a new selection strategy based on the concepts of clonal selection principle is used to maintain the balance between exploration and exploitation of the objective space. Simulation results demonstrate the effectiveness of the proposed approach in handling noise, plant uncertainties and the coupling effects of the cross-axis interactions.

This chapter presents a new framework for solving distributed control problems in a cooperative manner via the concept of dynamic team building. The distributed control problem is modeled as a set of sub-problems using a directed graph. Each node represents a sub-problem and each link represents the relationship between two nodes. A cooperative ensemble (CE) of agents is used to solve this problem. Agents are assigned to the nodes in the graph and each agent maintains a table of link relationship with all the other nodes of the problem. In the cooperative ensemble, each agent generates three sets of outputs iteratively based on the input variables it receives. They are, namely, the need for cooperation, the level of cooperation and the control directives. These outputs are used for dynamic team building within the cooperative ensemble. Agents within each team can issue a collaborative control directive and they take into account the mistakes of all the members in the team. In addition, each agent has a neuro-biologically inspired memory structure containing the addictive decaying value of all its previous errors and it is used to facilitate the dynamic update of the agent’s control parameters. The cooperative ensemble has been implemented in the form of distributed neural traffic signal controllers for the distributed real-time traffic signal control. It is evaluated in a large simulated traffic network together with several existing algorithms. Promising results have been obtained from the experiments. The cooperative ensemble is seen as a potential framework for similar distributed control problems.

Evolutionary computing (EC) techniques such as genetic algorithms (GA), genetic programming (GP), evolutionary programming (EP) and evolution strategies (ES) mimic nature through natural selection to perform complex optimisation processes. Grid-enabled environment provides suitable framework for EC techniques due to its computational and data capabilities. In addition, the semantic and knowledge Grids aid in the design search and exploration for multi-objective optimisation tasks. This chapter explores some problem solving environments such as Geodise (Grid-Enabled Optimisation Design Search for Engineering), FIPER (Federated Intelligent Product Environment), SOCER (Service-Oriented Concurrent Environment), DAME (Distributed Aircraft Maintenance Environment) and Globus toolkit to demonstrate how EC techniques can be performed more efficiently within a Grid environment. Service-oriented and autonomic computing features of Grid are discussed to highlight how EC algorithms can be published as services by service providers and used by service requestors dynamically. Grid computational steering and visualisation are features that can be used for real-time tuning of parameters and visual display of optimal solutions. This chapter demonstrates that grid-enabled evolutionary computing marks the future of optimisation techniques.